How SuperSport Lifted AOV by 16.6% With AI Upsells & Infinite Scroll

(Data measured from November 2025 to December 2025)

 

Meet SuperSport Vietnam

SuperSport stands among Southeast Asia’s most prominent sports and fitness retailers, with strong operations in both Vietnam and Thailand. Offering an expansive catalog that ranges from athletic footwear and performance apparel to fitness accessories and equipment, the brand serves a large and diverse customer base across multiple digital channels.

As its online presence continued to grow, SuperSport encountered a pivotal challenge: maintaining exceptional site performance while introducing AI-powered upselling and dynamic product discovery. The eCommerce team aimed to achieve a delicate balance between speed, personalization, and scalability, ensuring that every recommendation delivered value without compromising user experience or page efficiency.

To overcome these challenges, SuperSport partnered with Zotasell, deploying an integrated solution that combined Lazy Load Infinite Scroll with a custom AI recommendation engine. This collaboration led to measurable gains in page speed, conversion rate, and average order value (AOV), establishing a new standard for enterprise-level Shopify stores across the region.

 

The Challenge

1. Balancing Infinite Scroll Experience with Page Speed and Conversion

SuperSport set out to replicate the engaging and frictionless shopping experience popularized by regional eCommerce giants like Shopee. The brand aimed to integrate infinite scroll on its homepage – a feature that continuously loads products as users browse, to make product discovery effortless.

The vision was simple yet ambitious: allow shoppers to explore related products, trending items, and new arrivals in a single seamless flow, eliminating the need to reload pages or navigate through multiple categories. This design promised to increase upsell opportunities and encourage spontaneous purchases, both of which are essential drivers of AOV and long-term customer value.

However, during pilot implementation with third-party solutions, the reality proved far more complex. Infinite scroll, while visually dynamic, demanded intensive data processing. Each scroll triggered multiple API calls and database queries, which became increasingly heavy as the product catalog expanded to over 20,000 SKUs. The system began to show its limits:

• Page speed dropped significantly, especially on mobile networks.
• User experience declined, with visible content delays during browsing.
• Bounce rates rose sharply, as users abandoned sessions before reaching product recommendations.
• Conversion rates fell due to slower load times and disrupted shopping continuity.

In essence, what was meant to be a discovery-driven experience turned into a performance bottleneck. SuperSport’s team needed a solution that could preserve the visual engagement of infinite scroll while reducing load strain, ensuring that personalization and speed could coexist without compromise.

2. Difficulty Promoting New Products Efficiently

Beyond the performance issues, SuperSport faced another major challenge: the ineffective visibility of newly launched products.

As a retailer with frequent seasonal collections and new arrivals, SuperSport regularly refreshed its catalog to stay aligned with market trends. However, its existing upsell system relied heavily on historical behavioral data – past purchases, browsing activity, and cart additions. While this rule-based approach worked well for recommending established products, it created a critical blind spot. New items without any prior engagement data were rarely, if ever, surfaced to customers.

This limitation had clear business consequences.

• New arrivals lacked exposure, resulting in slower initial sales and reduced launch momentum.
• Upsell opportunities were lost, as the algorithm prioritized older SKUs with richer data histories.
• Click-through rates (CTR) declined, particularly for product recommendations that failed to feature the latest releases.
• Revenue diversification weakened, since the system kept promoting legacy products rather than emerging catalog entries.

The problem extended beyond algorithmic logic; it reflected a strategic constraint. SuperSport needed an upsell engine that could intelligently balance relevance and recency, dynamically identifying not just what customers had previously shown interest in, but also what was new, timely, and contextually appealing.

To solve this, the eCommerce and data teams aligned on a shared objective: to develop an AI-powered recommendation layer capable of adapting to real-time catalog changes. This solution would ensure every browsing session displayed an optimal mix of proven bestsellers and new arrivals, much like Shopee’s contextual “You May Like” carousel – a system that introduces trending products seamlessly during live sessions.

 

The Solution

To address SuperSport’s dual challenge of maintaining performance and improving product discovery, Zotasell engineered a comprehensive two-tier optimization strategy tailored to the brand’s enterprise-scale operations and vast product catalog.

1. Lazy Load Infinite Scroll: Combining Speed and Scalability

Zotasell fundamentally restructured SuperSport’s existing infinite scroll mechanism using a Lazy Load architecture, designed to balance dynamic content delivery with optimized system performance. Instead of loading all products simultaneously, the new architecture loads content progressively as users scroll through the page.

This adjustment significantly reduced initial page load time by up to 40%, creating a seamless and responsive browsing experience even when handling thousands of SKUs across multiple product categories.

Key advantages include:

Optimized Time-to-Interaction	Limited data processing per scroll ensured that users could begin browsing and interacting almost immediately.
Smooth, uninterrupted scrolling	The new experience maintained visual consistency and responsiveness across both desktop and mobile devices.
Higher engagement levels	Visitors viewed an average of 25% more products per session, reflecting improved session depth and interaction quality.
Sustained stability	Performance remained consistent even during large catalog updates or promotional campaigns.

By deploying this architecture, SuperSport successfully preserved the immersive nature of infinite scroll while eliminating the performance bottlenecks that had previously slowed navigation and reduced conversion potential.

2. Custom AI Algorithm for Dynamic Product Discovery

To resolve the visibility gap for newly launched items, Zotasell developed a custom AI-driven recommendation algorithm built around real-time user behavior and catalog updates. Unlike traditional static recommendation systems, this model dynamically analyzes browsing context, purchase intent, and catalog freshness to surface the most relevant and timely product suggestions within upsell modules.

The AI engine not only recognized products with high similarity and engagement potential but also intelligently prioritized new or trending releases, ensuring that fresh items were introduced to users throughout their shopping journey.

For example:
When a customer viewed a Basezip Long Compression Tee, Zotasell’s recommendation engine could display:

• Similar performance tops that match in fit, compression level, fabric technology, or price range.
  

• New arrivals within the same category, such as updated shoe models or complementary accessories.

Measured outcomes:

• 38% increase in new product visibility, as AI recommendations balanced relevance with novelty.
• 24% improvement in upsell CTR, as users interacted more frequently with timely, contextual suggestions.
• Steady AOV growth, driven by expanded discovery of new releases and cross-category exploration.

Through this dual-layered approach, pairing technical optimization with intelligent personalization, Zotasell enabled SuperSport to transform product discovery into a high-performance, revenue-driving experience, seamlessly integrating new product promotion into a fast, fluid, and data-driven user journey.

 

The Results

Zotasell’s solutions delivered measurable business and technical improvements within the first month of implementation, reinforcing the impact of AI-driven optimization on both performance and revenue growth.

1. Improved Site Performance

The Lazy Load Infinite Scroll system significantly strengthened browsing speed and session stability across SuperSport’s high-volume catalog.

• Page load time improved by 40%, reducing friction and helping pages render noticeably faster.
• Bounce rate dropped by 18%, as users stayed longer and scrolled deeper through product listings.
• Pages viewed per session increased by 25%, reflecting stronger browsing continuity and a smoother discovery flow.

These enhancements allowed SuperSport to maintain a visually rich shopping interface without compromising responsiveness or customer satisfaction, even with frequent catalog updates.

2. Upsell Revenue Impact and AOV Growth

Zotasell’s AI-powered recommendation engine delivered meaningful lifts in customer purchasing behavior from the very first week.

• Stronger Cart Engagement:
  Shoppers interacted more deeply with product recommendations, exploring a broader range of items during each session. This increased engagement set the foundation for larger carts and improved purchasing intent.
• 16.60% AOV Growth:
  More customers purchased complementary or higher-value items, leading to a measurable 16.60% increase in Average Order Value from the first week of deployment.
• 12.01% Upsell Rate (30 days):
  More than one in ten orders included at least one upsell item, demonstrating strong engagement with AI-powered recommendations at enterprise scale. Notably, the upsell rate reached 15.40% within the first 7 days, reflecting rapid adoption even during the initial rollout phase.
• High-value upsell contribution:
  The “Related Products” placement on the Product Page accounted for the majority of upsell-driven orders and value, confirming the precision of Zotasell’s recommendation logic in identifying high-intent moments during product exploration.

These improvements demonstrated that precise personalization, delivered at the right moment, can meaningfully expand cart size and elevate session value, even during the initial deployment phase.

3. Seamless Enterprise Integration

Zotasell’s architecture was designed to integrate effortlessly with SuperSport’s enterprise-scale Shopify setup.

• No complex code changes were required, ensuring smooth onboarding without operational disruption.
• Real-time inventory syncing guaranteed accuracy in product availability and pricing across all upsell placements.
• The system handled large product launches and peak-traffic windows with full stability, maintaining consistent UX quality throughout.

This seamless integration empowered SuperSport to scale its upsell strategy instantly while preserving the reliability and performance needed for an enterprise retail environment.

“Super easy to use and set up. The UI is clean and intuitive, and it doesn’t affect my store’s loading speed at all.”  – Super Sport

Why It Works

It is important to note that these results were achieved during SuperSport’s year-end sales period, supported by coordinated promotional efforts from the internal eCommerce and marketing teams. Seasonal demand, campaign execution, and operational readiness all played a meaningful role in driving higher traffic and purchase intent during this time.

At the same time, Zotasell’s contribution extended beyond passive support. Throughout the campaign window, the AI recommendation engine was continuously optimized based on live behavioral data, product availability, and real-time engagement signals. These algorithm updates allowed upsell placements to adapt dynamically to shifting shopper intent, ensuring that recommendations remained relevant even as traffic volume and browsing patterns fluctuated.

This combination of strong internal execution from SuperSport and adaptive AI optimization from Zotasell created a compounding effect. Rather than relying solely on seasonal uplift, SuperSport was able to convert increased demand into measurable upsell adoption and sustained AOV growth through intelligent, context-aware product recommendations.

1. Speed and Personalization in Harmony

The Lazy Load Infinite Scroll architecture allowed SuperSport to maintain exceptional page performance even with an extensive catalog of tens of thousands of SKUs. Personalization remained seamless and responsive, enabling real-time upsell delivery at scale with a smooth, uninterrupted user experience that kept shoppers engaged.

2. AI That Learns and Adapts 

Unlike conventional rule-based systems, Zotasell’s algorithm continuously learns from user intent, browsing behavior, and the timing of new product launches. Each recommendation becomes contextually relevant and commercially optimized, helping customers discover the right products while driving measurable business growth.

3. Data-Driven Scalability

Through Zotasell’s powerful analytics dashboard, SuperSport could closely monitor conversion uplift, evaluate product-level engagement, and refine its recommendation models with precision. This established a sustainable feedback loop of optimization, allowing the brand to evolve intelligently with every interaction.

Ultimately, Zotasell’s partnership empowered SuperSport to elevate upselling into an intelligent and performance-driven growth engine, demonstrating that AI-powered personalization can simultaneously increase revenue, enhance customer experience, and preserve optimal site speed.